HD FS 258 Lecture Notes - Lecture 28: Predictive Modelling, Odds Ratio
Document Summary
Correlation: x is related to y; y is related to x (-1. 00 to +1. 00) Predictive modeling: x at time 1 predicts y at time 2, using data to predict possible future outcomes. Mediation: x causes y through m, pertains to causation, there is a third variable that explains a predictive relationship between x and y. Correlation between # of sexual assaults & ice cream sales: both #s go up and down at same time. However, doesn"t cause the other variable: third variable: Risk factor: an event preceding an undesired outcome and increases the probability of that outcome. Predicting the probability of an adverse event. How big difference is between factual and counterfactual condition. Ex; how stark the color difference is between purple and clear. If clear is not so clear, odds ratio is small. Using others as a way to provide gains for self. Moms are more likely to participate in research. Parenting risk factors: low paternal involvement.